The set of a recording studio or the set of a movie may include numerous microphones that generate numerous audio recordings. For example, each person on the set may have a “bug microphone” that records his or her speech. Thus, these multiple audio recordings are generated in addition to the associated video recording. After recording, the audio recordings may be processed to generate audio channels for reproduction (e.g., playback for the enjoyment of an audience). The processing may generate surround sound audio channels (e.g., “5.1” audio channels), two channels for stereo or synthesized binaural sound, etc. Thus, the person watching a recorded video may perceive sound coming from the direction of the speakers. This audio processing may require an operator to enter the location of audio sources, for example, by tediously selecting points in the frames of the corresponding video.